Proceedings, International Snow Science Workshop, Banff, 2014

ANALYSIS OF THE WORKING OF DEFORMABLE BARRIERS USED TO MITIGATE THE HAZARD OF AVALANCHES CHANNELLED INTO A TRACK ZONE.TRIALS ON SITE IN WINTER 2013-2014

* Raviglione Massimo1, e Barberis Michela2

1 RocK and Snow Engineering, – 2 MountainS WorkinG sas, Tollegno – ITALY

ABSTRACT: The project of deformable structures capable of containing a channelled avalanche flow originated from the need to reduce hazard from sites where an avalanche flow coming from a large starting zone interferes, at a specific point, with a roadway or, in general, with a road communication. Under those conditions, risk mitigation provided by management procedures (such as road closing) may involve disruption and economic damage to the community while, in alternative, reducing hazard through structural works, such as works in active detachment zone, may involve high investment costs. The project presented here, even if it is an engineering protection work, does not reflect the traditional rules of structural intervention, but it involves the use of deformable barriers (net barriers installed perpendicularly to the avalanches flow) that intercept the avalanche, retain part of the flow and dissipate its energy both through the deformation of the structure and "the change in the sliding profile" of the flow thanks to the "check dams " function of the filled barrier. Starting from the theoretical outcome of the full-scale experimentation carried out in the avalanche area of "Rio Colombari" San Paolo Cervo, Piemont, Italy, located upstream of Provincial Road No. 513 " San Giovanni Oropa", with the present work we want to practically illustrate the functionality of the tested prototype, the cost-benefit analysis of the project, the effect of hazard reduction and, last but not least, the amount of maintenance required for the functional restoration of the works.

1. INTRODUCTION1 2. TEST SITE This paper presents the research activity and the 2.1 - Geographical Location results of the full scale tests performed to verify The area is part of the avalanche starting zone the behaviour of passive snow barrier installed called the Colombari Rio, located in the Southern along the sliding zone of channelled avalanches. Pennine Alps, about 26 km from the town of The analysis of the behaviour of the structure, (Piemonte - Italy), NNE direction, in Comune San carried out during winter 2013-2014, has allowed Paolo Cervo (latitude N 45,648° - longitude E us to assess the behaviour of such installations in 7,982° - altitude ≈ 1600 m). situations of dynamic impact and of stress due to the over-flow caused by events happened after the filling of the barrier.

After the observation of the avalanches in situ and Test of their effects on the territory, a back analysis with area 2D numerical modeling was carried out, starting from the nivo-meteorological data in the area, to assess the impact of dynamic loads (velocity and pressure of the avalanche flow) and the heights of the avalanche impacting the structure. Biella The study ends with an assessment of the structural damage and indications of the maintenance works which are necessary to the functional restoration of the barrier.

*Corresponding author address: Massimo Raviglione, Studio Tecnico Dott Ing Massimo Figure 1 – Geographical location of the test site Raviglione – RocK and Snow Engineering via Cesare Battisti 7 – 13818 Tollegno (BI) ITALY, e.mail [email protected]

1079 Proceedings, International Snow Science Workshop, Banff, 2014

Test area

Figure 4 – Graphic of HS per day

The most significative events (DH3gg ≥ 30 cm) in terms of increasing of snow height on three consecutive days of snowfall DH3gg, occurring Road SP 531 during the season 2013-2014, were:

• 17-22/11  DH3gg equal to ≈ 46 cm; • 24-26/12 DH3gg equal to ≈ 66 cm; • 16-19/01  DH3gg equal to ≈ 36 cm; • 28/01-05/02  DH3gg equal to ≈ 33 cm;

• 27/02-01/03  DH3gg equal to ≈ 55 cm. Figure 2 - Aerial view of the Colombari Rio avalanche site 2.3 - Avalanche features 2.2 - Features of snow-events The site is known as an avalanche area with a strong activity, in which the annual frequency of The snowfall that affected the area during the avalanches is equal to 1 - 2 events per year. winter season 2013-2014 deposited, in 23 significant cyclonic events (ie. event that is The starting zone consists of bowl-shaped areas of detachment characterized by medium - high associated with a snowfall HN ≥ 5 cm), an 2 inclinations (ψ = 35° - 40°) with classes of ground accumulated snow thickness equivalent to about 3 type 2 and 3 represented by coarse scree (NCH ≈ HS ≈ 4,80 m. 4 1,3 to 1,8 NFR or ≈ 2,6) and short grass The thickness of snow on the ground reached its interspersed with low bushes (NCH ≈ 1,8 to 2,4 or maximum in the first decade of March 2014 with NFR ≈ 2,6) with sliding zones in couloirs exceeding ≈ HS 2,50 m. 750 m.

Figure 3 – Graphic of accumulated HS and of HN in 3 days (DH3gg)

3 Glide factor – Suisse standard - UFAM WSL “Defence 2 Snow height from the ground, resulting from the data collected structures in avalanche starting zones” - Ed. 04-07 4 from the ARPA control unit - Bielmonte (UTM 32T E 428083 N Glide factor – French standard - NF P 95-304 “Equipements 5057024 altitude about 1480 m slm) de protection contre les avalanches – Filet paravalanches” - Ed. December 1992

1080 Proceedings, International Snow Science Workshop, Banff, 2014

Figure 7 – View of the detaching zone showing the the slab perimeter

As for the event covering the period 27/02-01/03 the avalanche activity was caused by a big snowfall in a short amount of time (DH3gg ≈ 55 cm e HN ≈ 55 cm ) which slipped on a sliding plane containing Sahara sand (degree of local danger level 4 – High for 1 day and level 3 – Considerable for 1 day).

Figure 5 – General view of the Colombari Rio avalanche site

The most significant avalanches occurred during the following periods: 24-26/12 and 27/02-01/03 which were characterized by high avalanche activity at all altitudes and exposures. In particular, for the event on 24-26 December, the avalanche activity was caused by big snowfall in a short period of time (DH3gg ≈ 66 cm and HN ≈ 66 cm) and a high degree of hazard (local danger level 4 – High for 3 consecutive days).

Figure 8 – General view of the Rio Colombari Figure 6 – View of the basin showing the test area basin showing the test area (photo dated 05/03/14)

1081 Proceedings, International Snow Science Workshop, Banff, 2014

new layer

snow with Sahara sand Figure 10 – View of the barrier installed along the avalanche sliding zone in Rio Colombari

The first type of stress occurred in the event avalanche on 24-26/12, in which the maximum old layer value of the avalanche flow was equal to the height of the barrier (hf ≈ hg), impacting the structure with a dynamic action distributed on the whole barrier surface. The second type occurred with the event avalanche on 27/02-01/03 in which the avalanche Figure 9 – Stratigraphic analysis dated 06/03/14 overflowed the barrier, impacting the structure already filled by the previous event, according to 3. EXPERIMENTATION AND TEST the "check dams" system. The tests performed in the 2013-2014 season The avalanches that affected the prototype are developed into two distinct steps: briefly summarized in the following photographs. Step A to assess the behavior, in real scale, of the barrier installed along the sliding zone at an altitude of about 1290 m asl (full scale test); Step B  to calculate (by back analysis of major avalanches) the dynamic stresses of the impact and the overpressure produced by the avalance overflowing the barrier.

3.1 - Step A  Full scale test The tests allowed to analyze the behaviour of the barrier subject to the dynamic impact of the Figure 11 – Detail of the avalanche overflowing the avalanche flow according to two types of stress: barrier in Rio Colombari (photo dated 12/01/14) the first due to the pressure perpendicular to the barrier (qd) while the second tangential (qf,r) and vertical for the over-pressure (qu + qa).

1082 Proceedings, International Snow Science Workshop, Banff, 2014

3.2 - Step B  Back Analysis The dynamic analysis of the avalanche events affecting the structure, and in particular the avalanches during the periods 24-26/12 and 27/02-01/03, was carried out by 2D analysis performed by AVAL-1D, a software specifically designed by the Institute WSL-SLF Davos (CH) to study the dynamics of avalanche flows. Starting from the snow and avalanches barrier characteristics of the site (calibration model), the simulation of the avalanches has allowed us to evaluate the dynamic characteristics of the flow both along the avalanche path (height, velocity and stopping point) and at the barrier (height and velocity). The following graphs show a summary of the results of the simulation for the two major avalanches that characterized the 2013-2014 winter season.

road runout zone SP513 (≈ 1200 m slm)

Figure 12 – View of the test site showing the barrier position (photo dated 06/02/14)

barrier

road Figure 14 – Height, velocity and pressure of the SP513 avalanche flow in the period 24-26/12 at the monitoring point

runout zone (≈ 1050 m slm)

Figure 13 – General view of the avalanche in Rio Colombari (photo dated 05/03/14)

1083 Proceedings, International Snow Science Workshop, Banff, 2014

1 2 2 pressure qf = ρ ⋅ ⋅⋅ vc = 76,8 kN/m 2 ffd

cd = 2 3 ρf = 0,3 t/m

Event dated 27/02-01/03 avalanche overflowing the barrier

height hf = 2,5 m

velocity vd = 15 m/s ρ 2 ⋅⋅⋅ δ fff senvh 2 pressure qu = = 14,8 kN/m lu 3 ρf = 0,45 t/m δ = 20°

lu = 5,85 m ρ 2 pressure qa = af gh = 12,5 kN/m 3 ρa = 0,5 t/m Figure 15 – Height, velocity and pressure of the g = 10 m/s2 avalanche flow in the period 27/02-01/03 at the pressure q = µ ⋅ = 4,4 kN/m2 monitoring point u,r qu µ = 0,3 4. RESULTS The tests carried out on the structures, the back 4.2 - Maintenance and functional restoration analysis study and the observations on the analysis behaviour of the barrier after the 2013-2014 winter The dynamic stresses, both direct due to the main season have allowed us to evaluate, in relation to flow of the avalanche and indirect due to the the dynamic characteristics of avalanches affecting "check dams" action, produced permanent plastic 5,6 the site, the pressures which acted on the deformations in the barrier caused by the structure as well as the maintenance and deformation of the brake system as well as by the functional restoration to be carried out at the end sliding of the interception ring panel on the main of the season. ropes.

4.1 - Stresses According to the data obtained from the back analysis, the stresses affecting the barrier for the two most significant avalanche events can be summarized as follows: Event dated 24-26/12 avalanche not overflowing the barrier height hf = 2,8 m velocity vf = 16,0 m/s

5 Suisse standard - WKF / AEAI (2005) Protection des objets contre les dangers naturels gravitationnels Figure 16 – View of the barrier after partial melting 6 ASTRA12 007 (2007) Actions d’avalanches sur les galleries of the snow (photo dated 05/06/14) de protection

1084 Proceedings, International Snow Science Workshop, Banff, 2014

6. REFERENCES AFNOR NF P 95-304 (1992) Equipements de protection contre les avalanches – Filet paravalanches Aineva (2005) Linee guida metodologiche per la perimetrazione delle aree esposte al pericolo valanghe Anena, Meteo France (2008) Eléments de nivologie McClung D., Schaerer P. (2005) The Avalanche Handbook Mears A. I. (1992) Snow-Avalanche hazard analysis for land- use planning ad engineering UFAM WSL SNV (2007) Defence structures in avalanche starting zones WKF / AEAI (2005) Protection des objets contre lesdangers naturels gravitationnels ASTRA 12 007 (2007) Actions d’avalanches sur les galleries de Figure 17 – View of the barrier after melting of the protection snow (photo dated 18/06/14)

For the functional restoration of the barrier, maintenance interventions are foreseen aimed at restoring the geometry and the pre-impact energy dissipation capacity of the structure. To sum up, the functional restoration foresees the implementation of the following major maintenance steps: a) restoration of the dissipation capacity of the barrier through the replacement of the deformed brake systems; b) restoration of the system geometry through the tensioning of the main upper/lower ropes and the re-positioning of the interception ring panel.

5. CONCLUSION As a consequence of the tests carried out and the results obtained it can be said that, concerning the structural and functional aspect related both to the dynamic action of the main avalanche flow and to the secondary action of over-flow, the prototype barrier reacted in the proper way and according to the function for which it was conceived and designed. The retention of the first snow mass allowed to reduce the stopping distance of the avalanche compared to the historically recorded limits, while the "check dams" function, even if tested at a single point, caused a loss of energy in the avalanche over-flow. Compared to active interventions in the detachment zone, this type of intervention guarantees a minor economic investment but requires, at the end of each winter season, suitable maintenance to restore the pre-impact level of performance.

1085